Characterization of signalling networks in ovarian cancer cells

Characterization of signalling networks in ovarian cancer cells

Norwegian University of Life Sciences

Faculty of Veterinary Medicine and Biosciences Department of Chemistry, Biotechnology and Food Science

Master Thesis 2014 60 credits

Lise-Lotte Flage-Larsen

For the degree in Master of Biotechnology (60 credits)

Characterization of signalling networks in ovarian cancer cells

Author

Lise-Lotte Flage-Larsen

Main Supervisor Kjetil Taskén

Co-Supervisor Lena Eroukhmanoff

Main Supervisor (NMBU) Tor Erling Lea

i

1 A CKNOWLEDGEMENT

First, I would like to thank my head supervisor Kjetil Taskén for giving me this opportunity to be a part of this great project and his group of outstanding scientists, and for the support and guidance throughout this master thesis.

I would like to extend a huge gratitude to my Co-supervisor Lena Eroukhmanoff, for her enormous support, guidance and enthusiastic spirit in the lab throughout the past year. Thank you for always being there and for never letting me feel bad, you are great scientist, colleague and friend.

Then I would like to thank my internal head supervisor at NMBU, Tor Erling Lea for your inputs and guidance throughout these last months.

Thanks to everyone at The Biotechnology centre and especially thanks to everyone in the Taskén group for their help and support.

I would also like to thank my family, for their support and complete understanding throughout this year, especially the last few months.

Finally yet importantly, I would like to thank my boyfriend for his understanding, love and never- ending support throughout this year, you “rocks my world”!

ii

2 A BSTRACT (E NGLISH )

Ovarian cancer is one of the most lethal gynaecological diseases worldwide, and in Norway approximately 500 women die from this disease every year. Women with ovarian cancer often experience vague symptoms that can be mistaken for much less severe diseases. This leads to late stage diagnosis of many patients, which is the main reason for why the overall five year survival rate is only around 40%. Accumulation of malignant ascites is frequently observed in ovarian cancer patients that have reached stage III and IV, and it is known to support tumour progression, by creating a tumour friendly microenvironment. Our understanding of malignant ascites and its effect on the intracellular signalling networks in ovarian cancer cells is still unclear. In this thesis we have tried to characterize intracellular phosphorylation pattern in ovarian cancer cell lines treated with ascites. We put particular focus on key proteins that are known to be involved in cell proliferation, cell survival and protein synthesis.

Our result show that several ovarian cancer cell lines; OVCAR-3, OVCAR-5, OVCAR-8, NCI/ADR-RES and SKOV-3, are suitable for the protocol that is established for these experiments.

Further experiments show that multiple intracellular proteins were phosphorylated upon treatment with ascites. Ovarian cancer cells that were treated with ascites from patient panel displayed with a different phosphorylation patterns, both between cell lines and patient samples. S-6 ribosomal protein exhibited a strong phosphorylation after SKOV-3 and OVCAR-8 cells were treated with ascites from most of the patients.

Ascites from any patient induced a distinct increase in phosphorylation between 3 and 20 minutes in both SKOV-3- and OVCAR-8 cells. STAT-3 phosphorylation seemed to be independent because of a low correlation with the phosphorylation of the other proteins. A strong phosphorylation was seen in STAT-3 after treatment with ascites from all the patient samples.

By using an IL-6R blocking antibody we showed that IL-6 was dominant when it comes to activation of JAK/STAT pathway and STAT-3 phosphorylation. There was no significant change in phosphorylation of other proteins when IL-6R was blocked and cells were treated with ascites.

Though, -the overall phosphorylation response amongst the other proteins were weak.

We also observed that STAT-1 displayed a moderate to strong phosphorylation in SKOV-3 cells when treated with some of the patient samples. Other proteins that were investigated, like AKT, MAPKAPK- 2 and MEK-1, displayed moderate phosphorylation after treatment with ascites from multiple patients in both OVCAR-8- and SKOV-3 cells.

iii

3 A BSTRACT (N ORWEGIAN )

Kreft i eggstokkene er en av de mest dødelige gynekologiske sykdommene i verden, og hvert år dør rundt 500 kvinner i Norge av denne sykdommen. Kvinner med eggstokkreft opplever ofte vage symptomer som ofte kan bli oversett til å være mindre alvorlige sykdommer. Dette fører ofte til en sein diagnose for mange pasienter. Dette er hovedårsaken til den den generelt lave

overlevningsraten på bare 40%. Akkumulering av ondartet ascites er ofte observert i pasienter med eggstokkreft som er diagnostisert med stadium III eller IV, og er kjent for å støtte progresjonen av svulst, ved å skape ett vennlig mikromiljø. Vår forståelse av ondartet ascites og dens effekt på det intracellulære signalnettverket i eggstokkreft celler er fremdeles noe uklart. I denne oppgaven har vi prøvd å karakterisere det intracellulære fosforylerings mønsteret i eggstokkreft celler som har blitt utsatt for ascites. Vi har spesielt fokusert på nøkkel proteiner som er kjent for å være involvert i forøkning og overlevelse av celler samt protein syntese.

Resultatene våre viste at flere cellelinjer var passende å bruke; OVCAR-3, OVCAR-5, OVCAR-8, NCI/ADR-RES and SKOV-3, med tanke på protokollen som er etablert for disse eksperimentene.

Eksperimentene viste at flere intracellulære proteiner ble fosforylert etter at de hadde blitt utsatt for ascites. Eggstokkreft celler som ble usatt for ascites fra pasient panelet viste forskjellige fosforyleringsmønstere, både mellom celle linjene og pasient prøvene. S-6 ribosomal protein viste en sterk fosforylering etter at SKOV-3- og OVCAR-8 celler ble utsatt for ascites fra mesteparten av pasientene.

Ascites fra enhver pasient induserte en bestemt økning i fosforyleringen på mellom 3 og 20 minutter i både SKOV-3- og OVCAR-8 celler. Fosforylering av STAT-3 viste seg å være uavhengig grunnet en lav korrelasjon med fosforyleringen til de andre proteinene som ble undersøkt. En sterk fosforylering av STAT-3 ble sett etter at celler ble utsatt for ascites fra alle pasientene.

Ved å bruke ett antistoff som blokkerte IL-6R kunne vi vise at IL-6 var dominant når det kom til å aktivere JAK/STAT og dermed STAT-3 fosforylering. Det var ingen signifikant endring i fosforyleringen av andre proteiner når IL-6R ble blokkert og cellene ble utsatt for ascites, men fosforyleringsnivået var generelt svakt blant proteinene som ble undersøkt.

Vi observerte også at STAT-1 viste moderat til sterk fosforylering i SKOV-3 celler etter de ble utsatt for ascites fra noen av pasientene. Andre proteiner som ble undersøkt, som for eksempel AKT, MAPKAPK-2 and MEK-1, viste moderat fosforylering etter at SKOV-3 og OVCAR-8 celler ble usatt for ascites fra flere av pasientene.

iv

4 A BBREVIATIONS

ACJJ American Joint Committee on Cancer AKT/PKD Protein kinase B

ARID1A AT rich interactive domain

ATX Autotaxin

BLC-2 B-cell lymphoma 2 gene BRCA 1 Breast cancer 1 and 2 BRCA 2 Breast cancer 1 and 2

BRAF V-raf murine sarcoma viral oncogene homolog B BSA Bovine serum albumin

CA-125 Cancer antigen 125

c-MYC V-myc avian myelocytomatosis viral oncogene

Cl^- Chloride

CT Computed tomography

CTNNB1 Beta-catenin DMSO Dimethyl sulfoxide DNA DeoxyriboNucleic Acid EGF Epidermal growth factor

EGFR Epidermal growth factor Receptor ERK Extracellular-signal-regulated kinases GEF Guanine exchange factor

FAB Fragment antigen binding FAK Focal adhesion kinase FC Fragment crystallisable FCS Fetal bovine serum FSC Forward side scatter

FIGO International Federation of Gynaecology and Obstetrics FITC Fluorescein isothiocyanate

GDP Guanosine diphospha

GRB2 Growth factor receptor-bound protein 2

v GTP Guanosine triphosphate

H3 protein Histone 3 protein

HER1-4 V-erb-b2 avian erythroblastic leukemia viral oncogene homolog 1,2,3 and 4 HGSC High grade serous carcinoma

HNF1 Homeobox A and B

HNPCC Hereditary nonpolyposis colorectal cancer (Lynch syndrome) IL-6 Interleukin 6

IL-8 Interleukin 8 IL-10 interleukin 10

IRS-1 insulin like growth factor 1

JAK Janus kinase

KRAS Kirsten rat sarcoma viral oncogene homolog LPA Lysophospatidic acid

LGSC Low grade serous carcinoma

MLL2 Histone-lysine N-methyltransferase 2 MLL3 Histone-lysine N-methyltransferase 3 MRI Magnetic Resonance Imaging

NCI-60 National Cancer Institute (human cancer panel 60)

OPG Osteoprotegerin

PARP Poly(ADP-Ribose)polymerase PBS Phosphate-buffered saline PerCP Peridinin chlorophyll protein PH Pleckstrin Homology domain PI3K Phosphoinositide 3-kinase

PIP2 Phosphatidylinositol 4,5-bisphosphate PIP3 Phosphatidylinositol (3,4,5)-trisphosphate PKD1 Phosphoinsositide dependent kinase 1 PKD2 Phosphoinsositide dependent kinase 2 PLA 1/2 Phospholipase A 1 and 2

PMT Photomultiplier tube PS Penicillin-Streptomycin

vi PTEN Phosphatase and tensin homolog SH2 Src (Schmidt-Ruppin A-2) homolog 2 SH3 Src (Schmidt-Ruppin A-2) homolog 3 SOS Son of Sevenless

RANKL Receptor activator of nuclear factor kappa-B ligand RANK Receptor activator of nuclear factor kappa-B RAS Rat sarcoma protein

RPMI Roswell Park Memorial Institute

SSC Side scatter

STAT Signal Transducer and Activator of Transcription TNFR Tumour necrosis factor receptor

TRAIL TNF-related apoptosis-inducing ligand TP53 Tumour protein 53

VEGF Vascular endothelial growth factor UPA Urokinase plasminogen activator

1

T ABLE OF C ONTENTS

5 Introduction... 3

5.1 Ovarian cancer ... 3

5.1.1 Epithelial ovarian cancer ... 4

5.1.2 Diagnosis ... 6

5.1.3 Treatment ... 7

5.1.4 Ascites accumulation – how and why ... 7

5.2 Key signalling pathways in ovarian cancer ... 9

5.3 Components in ascites ... 11

5.3.1 LPA (Lysophosphatidic acid) ... 11

5.3.2 Cytokines ... 12

5.3.3 EGF (Epidermal growth factor) and its signalling pathway ... 12

5.3.4 OPG (Osteoprotegerin) ... 13

5.3.5 VEGF (Vascular endothelial growth factor) ... 13

5.4 Ovarian cancer cell lines ... 14

5.5 Statistics... 16

6 Objectives ... 17

7 Materials ... 18

7.1 Cell culture ... 18

7.2 Flow cytometry ... 19

7.3 Western blot ... 22

8 Methods ... 24

8.1 Cell culture ... 24

8.1.1 Cell lines ... 25

8.1.2 Treatment ... 27

8.2 Flow cytometry ... 29

8.2.1 Spillover and compensation... 31

8.2.2 Fluorescent cell barcoding ... 33

8.2.3 Antibodies and staining ... 35

8.3 Cytobank – assessing results ... 38

8.4 Western blot ... 40

2

9 Results ... 44

9.1 Cold trypsinization of ovarian cancer cell lines ... 44

9.2 Ovarian cancer cell lines treated with EGF (epidermal growth factor) ... 44

9.3 Fluorescent cell barcoding ... 47

9.4 Ascites titration ... 47

9.5 Ascites-induced phosphorylation at different time points ... 50

9.6 Treatment of NCI/ADR-RES cells with ascites from patient #35 and #40 ... 51

9.7 Investigation of IL-6-induced phosphorylation in SKOV3 cells, with focus on STAT3 ... 55

9.8 Treatment of SKOV-3 cells with ascites samples from 20 patients ... 57

9.9 Treatment of OVCAR-8 with 18 ascites samples ... 60

10 Discussion ... 62

10.1 Cell lines ... 62

10.2 EGF stimulation ... 62

10.3 Ascites titration ... 63

10.4 Treatment time ... 63

10.5 Treatment of NCI/ADR-RES cells with two patient samples ... 64

10.6 IL-6R blocking antibody, effects on SKOV-3 cells ... 64

10.7 Ascites screen of SKOV-3- and OVCAR-8 cells ... 65

11 Conclusion ... 68

12 Further Work ... 70

13 References ... 71

14 Appendix I ... 73

14.1 OVCAR-5 ... 73

14.2 OVCAR-8 ... 75

14.3 NCI/ADR-RES ... 77

15 Appendix II ... 79

16 Appendix III ... 81

17 Appendix IV ... 83

18 Appendix V ... 85

3

5 I NTRODUCTION

5.1 O

Cancer of the ovaries is the fourth leading cause of cancer deaths among women worldwide. Every year 500 new cases of ovarian cancer are reported in Norway and the 5 year overall survival rate is approximately 40 % [2]. As shown in figure 1 the five year relative survival rate for women diagnosed with ovarian cancer is dependent on the spread of the disease at diagnosis. Women diagnosed at an early stage (localized) usually have a good prognosis. However, approximately 70 % of the patients are diagnosed when the cancer has metastasized to distant areas, within and beyond the abdominal cavity, and their prognosis is poor [2]. The low overall survival rate is mainly due to the high number of women that are diagnosed with advanced cancer.

With improved understanding of tumour, we can find new and better diagnostic tools that can help to detect ovarian cancer earlier, and thereby lead to better patient prognosis.

Figure 1 Five year relative survival rate dependent on the spread of the disease at diagnosis [3].

The main reason for the late diagnosis is the vague symptoms, making it difficult to detect cancer at an early stage. Some of the symptoms that women with ovarian cancer can suffer from are pelvic pain, abdominal bloating, vaginal bleeding and weight gain or loss. These symptoms can often be mistaken for much less severe diseases, thereby prolonging time to diagnosis. The exact causes of cancer in the ovaries are still unknown, but there are known risk factors associated with ovarian cancer. Some of them are; not given birth, not using oral contraceptives, obesity, increasing age and breast cancer due to the BRCA1 and BRCA2 mutations.

4

Prognosis and progression is dependent on where the tumour cells originate from. Ovarian cancer types can be divided in three main groups based on their morphological and biological behaviour [4]

(Figure 2A and B):

1) Surface epithelial carcinoma 2) Germ cell carcinoma

3) Stromal cell carcinomas

Approximately 90 % of ovarian cancer cases are classified as epithelial ovarian cancers, meaning that the tumour arises from the outer epithelial layer [4] (Figure 2B).

Figure 2 Abdominal cavity with focus on female reproductive organs (A) and locations for ovarian cancer origins (B), from [5]

5.1.1 Epithelial ovarian cancer

Epithelial ovarian cancer can be sub-divided into four main histological groups listed below [4, 6]:

1. Serous tumours

These types of tumours account for approximately 75% of all epithelial ovarian carcinomas worldwide [4]. There are two sub-groups within serous neoplasms due to their differences in molecular genetics; low grade serous carcinoma (LGSC) and high grade serous carcinoma (HGSC).

Low grade serous carcinoma is related to serous borderline tumours (benign tumours that develop into malignant tumours) and is frequently seen with mutations in BRAF (v-raf murine sarcoma viral oncogene homolog B) [6, 7]. Chemotherapy response and prognosis for patients diagnosed with LGSC are usually intermediate.

5

Unlike LGSC, HGSC is usually presented with mutations in TP53 (tumour protein 53) and BRCA (breast cancer 1 and 2) [6, 7]. This type of tumour is usually sensitive to chemotherapy, but patients with this diagnosis have a high probability for disease relapse and their prognoses are generally poor.

2. Mucinous tumours

This type of tumour is frequently confined to the ovaries and only appears in approximately 3-4% of ovarian cancer cases [4]. Patients have an elevated frequency of mutations in KRAS (kirsten rat sarcoma viral oncogene homolog) and HER2 (v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2) [6], and usually have a low sensitivity to chemotherapy treatment, but the prognosis is good due to surgery.

3. Endometrioid tumours

These tumours account for approximately 10 % of ovarian cancer cases [4]. One known risk factor for this tumour type is HNPCC (hereditary nonpolyposis colorectal cancer, also known as Lynch syndrome) and patients have an elevated frequency of mutations in PTEN

(phosphatase and tensin homolog), beta-catenin (CTNNB1) and ARID1A (AT rich interactive domain) [6]. Patients are usually diagnosed at early stages and respond well to

chemotherapy.

4. Clear cell tumours

Around 10 % of ovarian cancer cases are clear cell tumours [4]. Like endometrioid carcinoma, clear cell tumours are also associated with endometriosis. Patients usually have a favourable prognosis and mutations are frequently observed in HNF1 (HNF1 homeobox A and B) and ARID1A [6].

The different mutations that are found in each of the above described tumour types indicates that they are really different diseases which calls for individual treatment.

6 5.1.2 Diagnosis

FIGO (International Federation of Gynaecology and Obstetrics) and ACJJ (American Joint Committee on Cancer) are two similar staging systems for ovarian cancer that are used worldwide. These staging systems divide patient populations into groups depending on how the tumours have spread, and can reveal further progression, prognosis and treatment options (Table 1). FIGO and ACJJ both divide the ovarian cancer cases into four main groups followed by a number of subgroups and tumour grades.

The main four ovarian cancer staging groups are:

Table 1 FIGO ovarian cancer main staging groups [4].

Stage I Cancer cells are found in one or both ovaries Stage II Tumour involves one or both ovaries with pelvic extension or primary peritoneal cancer

Stage III Tumour involves one or both ovaries with confirmed spread to the peritoneum outside the pelvis and/or metastasised to retroperitoneal lymph nodes

Stage IV Distant metastasis excluding peritoneal metastasis

Diagnosis is usually set after the patient has gone through a physical examination followed by vaginal ultrasound, CT and MRI. Patients are also tested for elevated levels of CA-125 (cancer antigen 125) in serum. CA-125 is a tumour antigen that is found elevated in approximately 90 % of advanced stage epithelial ovarian carcinomas, but is only elevated in 50 % of early stage epithelial ovarian cancer patients [8]. There are also other cancers and benign conditions that can give elevated CA-125 levels so this marker alone cannot confirm diagnosis [9]. One of the main focus areas within ovarian cancer research is to find biomarkers that can detect cancer at an early stage with high specificity and sensitivity, making it possible to screen high risk patients or populations. Therefore it is important to continue studying tumour biology and its microenvironment so that one day it will be possible to improve patient prognosis and quality of life.

7 5.1.3 Treatment

As mentioned earlier the standard treatment for ovarian cancer is debulking surgery followed by chemotherapy (platinum- and taxane-based). There have been improvements when it comes to surgical techniques and chemotherapeutical regimes in the last decade, but the results have been modest when it comes to clinical outcome [10]. For patients that are diagnosed at advanced stages the probability of relapse is almost inevitable, and the development of chemoresistance is a significant problem when it comes to further treatment and clinical outcome [10].

However, the development of new analytical equipment with high throughput properties has given a better understanding of tumour biology. This is promising for new biologically targeted therapies that can give patients a more personalized treatment with a higher sensitivity, specificity and less side effects than traditional chemotherapy. The most promising targeted agents that already have passed randomized trials for ovarian cancer are anti-angiogenesis- (for example VEGF inhibition) and PARP inhibition agents (Inhibits PARP1 which is involved in DNA repair mechanism) [10, 11].

5.1.4 Ascites accumulation – how and why

Ascites is a term that is used when there is an accumulation of fluid in the peritoneal cavity. The reason for this accumulation can be a number of conditions, for example liver failure, tuberculosis and cancer [12]. The term malignant ascites is generally used when the peritoneal fluid tests positive for malignant cells and has an elevated level of lactate dehydrogenase (involved in tumour initiation and metabolism) [12, 13].

Malignant ascites is a complex mixture of soluble components (cytokines, chemokines and growth factors) and different cell types (free tumour cells, spheroids, mesothelial cells, fibroblasts, macrophages, white and red blood cells). Together they create a microenvironment that is crucial for further cancer progression, development of chemoresistance and recurrence of disease [12].

In a healthy human, the peritoneal cavity contains some fluid that works as a lubricant. The amount of this fluid is strictly regulated by secretion of small molecules from capillaries through the peritoneal membrane and reabsorption through lymphatic channels [12]. The main purpose of this fluid is to support organ mobility and easy transfer of solutes between adjacent organs and the peritoneum (Figure 3 A).

8

When tumour spreads to the abdominal cavity, there will be an elevated production of peritoneal fluid due to the disruption of the epithelial lining, increased leakiness of the tumour microvasculature, secretion from the tumour and obstruction of the lymphatic vessels [12]( Figure 3 A and B).

Figure 3 Normal peritoneal cavity (A), tumour metastases and ascites accumulation (B), modified from [12]

Most tumour cells grow rapidly and need an efficient supply of blood/oxygen (VEGF and other growth factors are involved in tumour angiogenesis). The newly formed blood vessels have a different architecture that supports further tumour progression. One of the main alterations is increased permeability so that cytokines, chemokines and free tumour cells can migrate into surrounding tissue and influence the microenvironment. Also tumour cells, associated stromal and immune cells secrete proteins at an elevated rate which leads to an increased volume of peritoneal fluid. In many cases the lymphatic channels that normally absorb the fluid from the peritoneal cavity and into the lymphatic circulation system can be blocked by tumour cells, restricting the reabsorption [12].

In this master thesis I have studied the effects of soluble components in malignant ascites, therefore cellular components will not be discussed further.

9

5.2 K

Cell growth is normally tightly controlled, but oncogenic (can cause or give rise malignant carcinoma) driver mutations in mitogenic signalling pathways combined with mutations in cellular DNA repair and/or cell cycle control mechanisms can lead to malignant transformation (cells obtains the properties of cancer). Mutations in different tumour-suppressor genes (genes that protect cells from developing cancer) and proto-oncogenes (can become an oncogene if mutated) can cause abnormal activation of signalling pathways leading to the loss of control when it comes to cell division. It is the effect of the downstream signalling events which decides tumour progression. Components found in malignant ascites can lead to the activation or deactivation of multiple signalling pathways in ovarian cancer cells. The three main signalling pathways are the PI3K (Phosphoinositide 3-kinase)-AKT/PKD (Protein kinase B), RAS-ERK (extracellular-signal-regulated kinases) and JAK (Janus kinase)/STAT (Signal Transducer and Activator of Transcription) pathways.

Figure 4 Key signalling pathways in ovarian cancer [14]

Activation of the RAS-ERK pathway

RAS-ERK pathway can be activated when a ligand binds to its cognate receptor which causes a conformational change that activates tyrosine kinases, leading to the phosphorylation of specific tyrosine residues in cytosolic domains. The adaptor protein GRB2 (growth factor receptor-bound protein 2) binds to the phosphorylated tyrosine residues. The SH3 (Src (Schmidt-Ruppin A-2) homolog 3) domain on GRB2 binds to SOS (Son of Sevenless) which exhibits GEF (guanine exchange

10

factor) activity [15]. SOS binds to the inactive RAS protein (small GTPase that function as a molecular switch. Can be activated and deactivated) and promotes dissociation from GDP (guanosine diphosphate) bound to RAS. This leads to the binding of GTP (guanosine triphosphate) to RAS and the dissociation of SOS [15]. RAS is now in its active form which leads to a phosphorylation cascade, starting with Raf. Activation of the RAS-ERK pathway supports cell cycle progression and proliferation, but it can also activate the AKT pathway [14] (Figure 4).

Activation of the AKT pathway

AKT can be activated via RAS, for example by ligand binding to insulin receptors or insulin like growth factor 1 (IRS-1) receptors. When a ligand binds to the receptor, a conformational change occurs which leads to the recruitment of IRS-1(insulin receptor substrate 1) that binds to Src homolog 2 (SH2) domains on the receptor. Specific tyrosine residues on IRS-1 are phosphorylated which results in downstream signalling transduction by further phosphorylation of PI3K. PI3K is now activated and can generate PIP3 (Phosphatidylinositol (3,4,5)-trisphosphate) from PIP2 (Phosphatidylinositol 4,5- bisphosphate) [14].

AKT exhibits kinase activity, and when PH domain (Pleckstrin Homology domain) in AKT binds to PIP3, PDK1 (phosphoinsositide dependent kinase 1) and PDK2 (phosphoinsositide dependent kinase 2) are activated in two steps and AKT is phosphorylated [14]. This results in a fully activated AKT, which leads to a cascade of intracellular signalling. The major effects of the activation of AKT pathways in ovarian cancer are seen in figure 4.

Activation of the JAK/STAT pathway

When ligand binds to its respective receptor, a conformational change occurs, leading to the recruitment of JAK1/2. Trans-phosphorylation of JAKs (tyrosine residues in the cytosolic domains) results in the phosphorylation of recruited STATs [16]. Once STATs are phosphorylated they dimerise and translocate to the nucleus were they initiate transcription of multiple genes that promote cell survival/growth and differentiation.

11

5.3 C

Malignant ascites seems to play an essential role in tumour progression by creating a “tumour friendly” microenvironment [12, 17]. This fluid contains soluble components like cytokines, chemokines and growth factors, some of which are known to be overexpressed in ovarian cancer [12]. These components can affect key signalling pathways in ovarian cancer cells, by supporting and promoting cell cycle progression, cell proliferation, cell survival, protein synthesis and cell growth (Figure 4).

5.3.1 LPA (Lysophosphatidic acid)

LPA is one of the most known extracellular phospholipids that can induce a diversity of cellular responses, and are generated by autotaxin (ATX) and phospholipase A1/A2 (PLA1/2). The biological functions of LPA are mediated by G-coupled receptors (Gq, Gi and G12/13) leading to the activation of multiple signalling pathways (Figure 5). Activation of the PI3K-AKT and the RAS-ERK pathways by LPA leads to cell survival by supressing apoptosis and cell proliferation.

Figure 5 LPA activation of down-stream signalling pathways [18]

LPA triggers cellular responses that effect transcriptional regulation of multiple growth factors, like VEGF, UPA (urokinase plasminogen activator), IL-6 and IL-8 [12, 18]. This response can lead to increased endothelial permeability and inhibition of gap junction communication between adjacent cells [18], and thus support production and accumulation of ascites fluid. In ovarian cancer patients LPA levels are elevated in both serum and ascites, which correlates to a poor prognosis [12].

12 5.3.2 Cytokines

Cytokines are small polypeptides produced by cells, that individually or through cross talk with other cytokines or growth factors affect cancer progression through different mechanisms [16, 17]. Recent studies profiling malignant ascites have revealed elevated levels of multiple cytokines, including IL-6, IL-8 and IL-10 [12, 17] in ovarian cancer patients. IL-6 is regarded as a major contributor in tumourigenesis and leads to the activation of STAT3 which is known to initiate transcription of VEGF, BCL-2, c-MYC and Cyclin D1, and more. Cytokine activation of signalling pathways can promote cell survival, proliferation, migration/invasion angiogenesis and in some cases the development of chemoresistance [12, 17].

5.3.3 EGF (Epidermal growth factor) and its signalling pathway

Epidermal growth factor is a mitogenic stimulus for many cancer types. As seen in figure 6, EGF activation of its receptor leads to the phosphorylation of multiple intracellular proteins, for example AKT and STAT3. The EGFR (epithelial growth factor receptor) family consists of several receptors, HER2 (human epithelial receptor 2) , HER3 (human epithelial receptor3) and HER4 (human epithelial receptor 4), in addition to EGFR (human epithelial receptor 1) [19].

Figure 6 EGF signalling pathways

One of the known ligands that can bind and activate these receptors is EGF. However, HER2 and HER3 are thought to be co-receptors, because they do not bind ligands. EGFR is overexpressed in many cancers, due to mutations in genes coding for EGFR or in proteins that support EGFR activation and/or deactivation. Constant activation of EGFR results in uncontrolled cell division, and thereby supports tumourigenesis.

13 5.3.4 OPG (Osteoprotegerin)

OPG is a secreted member of the TNFR (tumour necrosis factor receptor) superfamily and it is known to regulate the homeostasis of bone re-modelling by preventing RANKL (Receptor activator of nuclear factor kappa-B ligand) binding to its receptor RANK (Receptor activator of nuclear factor kappa-B) (Figure 7).

Recent studies have revealed that OPG phosphorylates FAK (focal adhesion kinase) which results in AKT activation, supports proliferation and inhibits TRAIL-induced apoptosis in ovarian cancer cells [20, 21]. It has also been suggested that OPG works as a decoy receptor by binding to TRAIL with low affinity and inhibit TRAIL-induced apoptosis in other types of cancers, thus evading cell death (Figure 6). Recent research has shown that elevated levels of OPG are associated with a shorter progression- free survival [20]

Figure 7 OPG inhibitions of RANKL and TRAIL [21].

5.3.5 VEGF (Vascular endothelial growth factor)

VEGF is a known growth factor and its expression is associated with the advanced stages of ovarian cancer [12]. There are multiple members in the VEGF family, including VEGF-A, VEGF-B, VEGF- C, VEGF-D and VEGF-E that all derive from the same gene (VEGF) by alternative splicing. Signalling transduction is mediated by tyrosine kinase receptors, VEGFR-1, VEGFR-2 and VEGFR-3 (Figure 8).

Activation of VEGF pathways is known to increase angiogenesis, vasculogenesis (formation of new blood vessels, no pre-existing blood vessels) and lymphogiogenesis (formation of lymphatic vessels from pre-existing lymphatic vessels), thereby supporting tumourigenesis [22].

14

Overexpression of these receptors has been associated with the development of malignant ascites and poorer prognosis of ovarian cancer patients [12].

One of the mechanisms of VEGF activation is the down-regulation of the tight junction protein claudin 5 that increases the permeability of the peritoneal membrane by inducing angiogenesis. This leads to formation of new blood vessels with increased permeability [23]. Tight junctions are important to regulate the flow of solutes and water between epithelial and endothelial cell sheets.

5.4 O

Cell lines are obtained from tumour cells in patients diagnosed with ovarian cancer. These cell lines have been characterized extensively to ensure that results can be reproduced and compared between research communities.

The NCI-60 (National Cancer Institute) human cancer cell line panel is one of the most studied panels worldwide and it has been intensely investigated so that research data can be compared between laboratories around the world [24, 25]. There are seven ovarian cancer cell lines in this panel, all with specific mutation signatures. Each of the cell lines represents individual tumours from which they were originally derived and will therefore exhibit somewhat different intracellular phosphorylation patterns when cells are treated.

Figure 8 VEGF-A, B, C, D and E binding to its respective receptor which results in promotion of cell growth.

15

Figure 9 The 20 most frequently mutated genes found in ovarian cancer cell lines [26].

The most frequently observed mutation in ovarian cancer is TP53 which is an important tumour suppressor gene that normally contributes to the cellular and genetic stability of the cell [27] (Figure 9). A mutation in this gene might lead to loss of control in the cell division machinery, and tumours or other diseases may arise. Two other genes that are frequently mutated in ovarian cancer are MLL2 (Histone-lysine N-methyltransferase 2) and MLL3 (Histone-lysine N-methyltransferase 3) and these genes code for proteins involved with the methylation of the H3 protein (Histone 3). Mutations in H3 can lead to increased transcription of mitogenic genes [23, 24]. KRAS is another gene that is found mutated in multiple ovarian cancer cases and is an important protein when it comes to cell signalling [28]. These mutational differences are one of the reasons for why ovarian cancer is a very complex and heterogeneous disease. The ovarian cancer cell lines used in our experiments are listed in table 2.

Table 2 List of some known mutations in census genes in the different ovarian cancer cell lines from the NCI-60 panel, information is obtained from the cosmic - cell line project [26]

IGROV-1 SKOV-3 NCI/ADR-RES OVCAR-3 OVCAR-4 OVCAR-5 OVCAR-8

TP53 ARID1A ERBB2 (HER2) PIK3R1 TP53 KRAS CTNNB1

MLH1 FBXW7 TP53 TP53 BRCA2 APC

PIK3CA TP53 KRAS

PTEN APC ERBB2 (HER2)

16

5.5 S

Results in this thesis are presented as heatmaps, histograms, dot plots, bar diagrams and line diagrams. In heatmaps and histograms the results are calculated with a cloud based software (Cytobank) where elevated signals are displayed as colour intensities in proportion to a baseline signal, giving a visual overview of analysed data.

Most of our experiments are repeated once (n=1) or multiple times (n=1-5), making it possible to calculate the average (A)

∑

Were n, and x are the total number of values and each value, respectively, and the variations between the experiments (standard deviation, σ)

√ (∑( )

)

where N, n, x and µ are the total number of values, one specific value, each value and the mean of values, respectively.

17

6 O BJECTIVES

Ovarian cancer is the fifth leading cause of cancer deaths worldwide. A common complication with this disease is the accumulation of ascites. Ascites creates a tumour friendly microenvironment that further supports cancer progression. To better understand the tumour microenvironment we wanted to study the intracellular mitogenic signalling networks involved in multiple adherent ovarian cancer cell lines treated with malignant ascites, looking at phosphorylation patterns. The work presented in this master thesis is part of a larger project funded by the Norwegian Cancer Society with a Postdoc fellowship to Dr. Lena Eroukhmanoff to study the effects of the tumour microenvironment in ovarian cancer. As a part of this project, I characterized mitogenic signalling that mirrors the tumour microenvironment in several ovarian cancer cell lines in the presence of ascites from different patients. The objectives of my work were to:

1) Find good cell line candidates for our experiments;

2) Examine whether intracellular phosphorylation is affected by the cold trypsinization protocol, by parallel flow cytometry and Western blot experiments;

3) Ensure that 3D fluorescent cell barcoding with Alexa Fluor 488, Pacific blue and Pacific orange can be used in our experiments;

4) Study IL-6-induced intracellular phosphorylation of proteins, with focus on STAT3; and

5) Characterize intracellular phosphorylation signatures in two cell lines (SKOV-3 and OVCAR-8) treated with multiple patient samples of malignant ascites

18

7 M ATERIALS

In this section we list the chemicals, materials, equipment and instrumentation used during the execution of this thesis.

7.1 C

Table 3 Chemicals and dilutions used when working with cells. Producer full name: LT= Life technologies, SA=Sigma-Aldrich, BC= The Biotechnology Centre of Oslo, BD= BD Bioscience

ID Chemicals Producer / Cat# Dilutions

A1 1640 RPMI medium (1x) + GlutaMAX LT / 61870-010

A2 1640 RPMI medium + PS 500 mL 1640 RPMI medium (A1)

+ 5 mL PS (A7)

A3 1640 RPMI medium + FCS + PS 500 mL 1640 RPMI medium (A1)

+ 50 mL FCS (A6) and 5 mL PS (A7)

A4 1640 RPMI medium + PS + FCS + DMSO 500 mL 1640 RPMI medium (A1)

+ 5 mL PS (A7) + 50mL FCS (A6) DMSO (A8)

A5 TrypLE Express (1x) LT / 12605-010 A6 Trypsin (10xx) 2,5% LT / 15090-046

A7 FCS (Fetal bovine serum) LT / 26140-079

A8 PS (Penicillin-Streptomycin) LT / 10177012

A9 DMSO (Dimethyl sulfoxide) SA / 41640-100ML

A10 PBS BC

A11 Ethanol Kemetyl

A12 Barrycidal 36 Interchem hygiene GmbH

A13 Relu+On Virkon DuPont

A14 Fix buffer I BD / 557870

A15 EGF SA / E9644

A16 IL-6 Blocking antibody MAB227

A17 Alexa 647 CD3 LT / MHCD0320

19

Table 4 Cell lines, equipment and instruments/software used when working with cell culture.

7.2 F

Table 5 Chemicals used when performing flow cytometry. Producers full name; BD=BD Bioscience, SA=Sigma-Aldrich, BC=The Biotechnology Centre of Oslo

ID Chemicals Producer Dilutions

B1 Flow wash buffer BC 400 mL PBS (A10) + 5mL FCS (A7) + 10 % NaAzid (B2)

B2 10 % NaAzid BC

B3 FACS Clean BD / 340345

B4 BD FACS Shutdown solution BD / 334224

B5 BD FACS Flow BD / 342003

B6 BD Sheat solution BD / 336911 B7 Sodium hypochlorite solution SA / 71696-5L B8 Perm buffer III BD / 558050 B9 Negative beads BD / 51-90-9001291 B10 Positive beads BD / 51-90-9001229

B11 dH2O BC

ID Cell lines Producer

I SKOV-3 National Cancer Institute

II NCI/ADR-RES National Cancer Institute

III OVCAR-3 National Cancer Institute

IV OVCAR-4 National Cancer Institute

V OVCAR-5 National Cancer Institute

VI OVCAR-8 National Cancer Institute

ID Equipment Producer / Cat#

E1 50 mL sterile tubes (PP) Sarstedt / 62.547.254 E2 15 mL sterile tubes (PP) Sarstedt / 62.554001 E3 6 welled plate nuclon delta surface Thermo Scientific / 140675 E4 Costar 96 welled V bottom plate Sigma-Aldrich / CLS3894 E5 Nunc Cell Culture Treated EasYFlasks 75 cm² Thermo Scientific / 156499 E6 Costar Cell culture flasks 162 cm² Sigma-Aldrich / 3151 E7 Cryotube sterile vials Thermo Scientific / 377267

ID Instrument Software

F1 Leica DMIL microscope -

F2 Allegra X-22R centrifuge -

F3 Eppendorf centrifuge 5415R -

20

Table 6 Conjugated antibodies and their dilutions. Producer full name: BD=BD Bioscience, CST= Cell signalling technologies, SC=Santa Cruz

Conjugated antibodies Concentration Antibody FACS buffer Producer / Cat#

IgG kappa 5 µl 20 µl BD / 557783

Akt/PKB (pS473) 2 µl 23 µl CST / 4075

MEK1 5 µl 20 µl BD / 560043

NFkB (pS529) 1,5 µl 23,5 µl BD / 558422

NFkB (pS539) 2,5 µl 22,5 µl CST / 4887

STAT 3 4 µl 21 µl BD / 557815

MAPKAPK-2 1 µl 24 µl CST / 4320

Stat6 5 µl 20 µl BD / 612601

Rb 2 µl 23 µl BD / 558590

S6rp 1,5 µl 23,5 µl CST / 4851

44/42 MAPK 1 µl 124 µl CST / 4375

p38 MAPK 1 µl 49 µl CST / 4552

STAT 1 (pY701) 5 µl 20 µl BD / 512597

STAT 1 (pY701) 5 µl 20 µl BD / 560190

AKT (pT308) 5 µl 20 µl CST / 3375

STAT5 5 µl 20 µl BD / 612597

ATF-2 1 µl 24 µl SC / SC-8398

Histone H3 1,5 µl 48,5 µl CST / 9716

SAPK/JNK 2 µl 23 µl CST / 9257

Table 7 Unconjugated antibodies and their dilutions. Producer full name: BD=BD Bioscience, CST= Cell signalling technologies, SC=Santa Cruz, AC=Abcam, I=Invitrogen

Unconjugated antibodies Amount antibody Amount FACS buffer Producer / Cat#

PLCg-1 (pT783) 1 µl 250 µl CST / 2821

b-catenin (pS45) 1 µl 250 µl CST / 9564

Gsk3 beta (pS9) 1 µl 125 µl CST / 9336S

Gsk3 alfa (pS21) 1 µl 125 µl CST / 9316

Vasp (pS157) 1 µl 500 µl CST / 3111

Vimentin (pS56) 1 µl 125 µl AC / Ab52942

Jak-2 1 µl 200 µl CST / 8082S

CaMKII (pT286) 1 µl 125 µl CST / 3361

Cdc-2 (pT161) 1 µl 125 µl CST / 9114

HSP27 (pS82) 1 µl 125 µl CST / 2405

PKD/PKCµ 1 µl 500 µl CST / 2054

Vasp (pS239) 1 µl 500 µl CST / 3114

Vav (pY174) 1 µl 500 µl SC / SC16408-R

Goat anti rabbit A647 1 µl 8000 µl I / A21245

21

Table 8 Dyes used when performing flow cytometry. Producers full name: LT=Life technologies.

Dye name Producer Concentration Dilutions

Alexa Fluor 488 LT / A-20100 Stock

0 2 µL stock /dye stock) and 998 µL DMSO 1 100 µL (concentration 0) and 900 µL DMSO 2 100 µL (concentration 1) and 250 µL DMSO 3 100 µL (concentration 2) and 300 µL DMSO 4 50 µL (concentration 3) and 550 µL DMSO Pacific Blue LT / P-10163 0 2 µL stock /dye stock) and 498 µL DMSO

1 100 µL (concentration 0) and 900 µL DMSO 2 100 µL (concentration 1) and 300 µL DMSO 3 100 µL (concentration 2) and 300 µL DMSO 4 44 µL (concentration 3) and 355 µL DMSO Pacific Orange LT / P30253 0 10 µL stock /dye stock) and 90 µL DMSO

1 90 µL (concentration 1) and 360 µL DMSO

2 125 µL (concentration20) and 325 µL DMSO

3 75 µL (concentration 0) and 325 µL DMSO 4 83 µL (concentration 0) and 917 µL DMSO

Table 9 Visual observation of patient samples (ascites).

Patient

samples Visual observations Patient samples

Visual observations

# 23 Yellow and clear # 36 Yellow and clear

# 25 Weakly orange

and clear # 37 Yellow and clear

# 27 Weakly yellow

and clear # 38 Weakly yellow and clear

# 28 Yellow and clear # 39 Yellow with

tissue lumps

# 29 Yellow and clear # 40 Orange and clear

# 30 Yellow and clear # 41 Orange and clear

# 31 Yellow and clear # 42 Orange and clear

# 32 Yellow and clear # 43 Orange, unclear and with tissue lumps

# 33 Yellow and clear # 44 Orange/brown and clear

# 35 Yellow and clear # 45 Yellow and clear

22

Table 10 Cell lines, equipment and instruments/software used when performing flow cytometry.

7.3 W

Table 11 Chemicals used when performing Western blot. Producer full name; JIR=Jackson ImmunoResearch, BC=The Biotechnology Centre of Oslo, SA=Sigma Aldrich, TH=Thermo Scientific, AGFA=AGFA

ID Chemicals Producer Dilutions

C1 Lysis buffer BC 150mM NaCl + 50 mM Tris 8.0 + 1% Trito X-100

C2 Protease inhibitor 1 tablet + 1 ml dH20 = 50x solution. Further

diluted 50x in lysis buffer C3 Loading buffer (SDS 3x) BC

C4 Loading buffer (SDS 1x) BC

C5 SDS running buffer (10x) BC (150 g Tris/HCL + 720g glycine + 50g SDS) diluted to 5 litre with dH2O

C6 Towbin buffer BC (15g Tris/HCL + 1000mL methanol + 72g glycine)

diluted to 5 litre with dH2O, pH adjusted to 8,3

C7 TBS-T (10x) BC (60,5g Tris/HCL + 438,5g NaCl + 20 mL Tween 20)

diluted to 5 litre with dH2O

C8 TBS-T (1x) BC 100 mL of 10x TBS-T + 900 mL dH2O

C9 Ponceau S solution (0,1%) SA / P7170-1L

C10 Methanol VWR / 00288

C11 pSTAT-3 (Rabbit) CST / 9131 C12 pAKT (pT308) (Rabbit) CST / C13 Vinculin (mouse) SA / V9131

C14 BSA SA / A6003-25G

ID Equipment Producer / Cat#

E2 15 mL sterile tubes (PP) Sarstedt / 62.554001 E8 Costar 96 well V bottom plates Sigma-Aldrich / CLS3894

ID Instruments Software

F2 Allegra X-22R centrifuge - F3 Eppendorf centrifuge 5415R - F4 Eppendorf centrifuge 5810 -

F5 Vortex V1 -

F6 BD FACSCanto II FACSDiva

F7 BD LSRFortessa FACSDiva

23

C15 NaN3 SA / 26628-22-8

C16 Milk powder Tine

C17 5% milk solution BC 0,5 g milk powder(C16) to 10mL 1x TBS-T C18 Secondary antibody (mouse) JIR / 115035-146

C19 Secondary antibody (rabbit) JIR / 111-035-144 C20 Super Signal West Dura ext. TS / 34076 C20 G354 Rapid fixer AGFA / 2828Q C21 G153 A Developer AGFA / HT536 C22 G153 B Developer AGFA /HT536

C23 Gel Citerion with 10 % Tris-HCL polyacrylamide (18 wells)

C24 Ladder (standard)

Table 12 Equipment and instruments/software used when performing Western blot.

ID Equipment Producer / Cat#

E10 Development cassette

E11 Amersham hyperfilm MP GM healthcare

E12 Electrophorese chamber, criterion BIO-RAD E13 Transfer chamber, criterion blotter (strings) BIO-RAD

E14 Transfer membrane Immobilon Millipore Corporation / IPVH00010

ID Instruments Producer / Software

F8 CURIX 60 Developer AGFA

F9 Shaker platform STR8 Stuart Scientific

F10 Heater Dri-block DB2D Techne

F11 Electrophorese PowerPac HC BIO-RAD

24

8 M ETHODS

In this section we list the employed methodology. We used protocols modified by scientists in the Taskén group for all experiments. Some of these protocols are still under development.

8.1 C

Adherent cells can be grown in culture flasks with a surface that is slightly hydrophilic (pulls cells towards the plastic binding surface). Different plasma membrane proteins bind the cell to the flask surface and to other cells as they expand.

To provide a good environment for the cells it is important to carefully check them in the microscope, provide them with nutrients and to maintain conditions of right pH, stable CO2- and humidity levels.

Culturing conditions may vary from cell line to cell line and it is therefore important to check this before a new cell line is used. It is also important to split cells regularly to avoid mutations and cell death.

As illustrated by the standard growth curve for adherent cells (figure 10), they go through different growth stages. In our experiments, we wanted to examine cells in their log phase. If cells were to move beyond log phase they would experience stress and activate pathways that normally are inhibited during the log phase. This could lead to increased phosphorylation of some proteins and give a false baseline in our experiments.

Figure 10 General cell growth curve.

25 8.1.1 Cell lines

In our experiments we have used multiple cell lines which were stored at -150 degrees Celsius until used. When stock cell lines were thawed and expanded a personal stock was frozen down and used in later experiments.

8.1.1.1

1) Remove FCS medium 2) Rinse with 9 mL PBS (A10)

3) Add 2-4 mL of trypsin (A5). Trypsin need to be in contact with the entire surface were cells are attached to the- flask bottom

4) Remove trypsin

5) Incubate at 37°C with a CO2 level at 5 % until they detach – confirm by looking at the cells in the microscope (Leica DMIL – F1)

6) Add FCS medium (A3) to quench reaction, volume is dependent on split ratio 7) Distribute medium into new bottles (E6) or a tubes (E1, E2 - for cell counting)

8) Wash with FCS medium (A3) and distribute cell suspension into one or all the flasks/tubes 9) For 162cm³ flasks the total medium volume should be 25 mL, and for 75cm³ flasks the total

volume should be 12 mL

8.1.1.2 Cell line start-up and freezing

- Thaw cells fast in water bath and immediately transfer cells to preheated 11 mL FCS medium (A3) in culturing flasks 75cm³ (E5) and incubate at 37°C with a CO2 level at 5 %.

- Next day, remove medium from flask and add 12 mL new preheated FCS medium (A3) - Change medium every day until cells have reached approximately 80 % confluence

- Passage cells 1:2 and expand further to approximately 80 % confluency, then passage cells in flask into multiple flasks (E6) (See protocol for passaging cell lines, above)

26

- Once cells have reached about 80 % confluency they are ready to be trypsinazed and frozen down

- Trypsinize cells with 1x trypsin (A5) until all cells detach from surface (verify in Leica DMIL microscope, F1)

- Transfer cell suspension to 50 mL tubes (E1) and spin down for 7 minutes at 4°C at 200rcf, Allgera centrifuge (F2)

- Remove medium and add new FCS medium with DMSO (A4), re-suspend cells by carefully pipetting

- Add 1 mL to 3-4 2,5 mL cryo tubes (E7) pr. 162 cm³ flask and put on ice and transfer tubes to freezer (-150°C) for storage until needed

8.1.1.3 Setting up experiments with 6 well plates

In our experiments we have used 6 well plates (E3) with approximately 150 000 cells per well. It is important that counting is performed correctly in a Bürker chamber. It is recommended to count at least two individual times and calculate the average.

- Perform step 1-5 listed in the protocol for passaging cell lines, above - Add FCS medium (A3) to quench reaction, volume is not important - Transfer cell suspension into sterile tubes (E1 or E2)

- Spin down tube at 200rcf for 7 minutes, Allgera centrifuge (F2) - Remove supernatant

- Re-suspend cell pellet in 2-12 mL FCS medium (A3) depending on the pellet size - Mix 100 µL PBS and 100 µL cells in a 1.5 mL sterile Eppendorf tube

- Add 12,5 µL to counting chip

- Count cells in microscope (Leica DMIL – F1) and follow the guidelines, figure 10 - Repeat the two last points

- Dilute cell suspension so that there is 150 000 cells per mL in FCS medium (A3). Take into consideration how many 6 well plates (E3) that are used in the experiment

27

- Add 1 mL into each well and incubate at 37° with a CO2 concentration at 5 % overnight - Remove medium and wash with FCS free medium (A2)

- Add 1 mL FCS free medium (A2) into each well (starve the cells), and incubate at 37° with a CO2 concentration at 5 % overnight

- Before starting experiment, check the cells in the microscope, Leica DMIL (F1), and make sure they are healthy and evenly distributed. Cells in plates should now be ready for stimulation.

8.1.2 Treatment

8.1.2.1 Treatment medium

In our experiments, we have treated cells with EGF or with ascites from multiple patients. EGF was used to investigate whether intracellular phosphorylation was affected by the established cold trypsinization protocol (EGF is known to trigger cellular responses in most cell lines). We used a concentration of 100 ng/mL in our experiments.

Ascites samples were collected at Haukeland University Hospital from patients with ovarian cancer in stage III and IV. Samples were collected at the time of surgery and before chemotherapy. The Regional Ethics Committee approved the study and all patients consented. When received, samples were logged and spun down to remove cells (tumour and immune cells), then stored at -80 °C until use.

8.1.2.2 Treatment protocol

- For each well treated, approximately 260-280 µL ascites is needed

- Thaw ascites samples in room temperature, put immediately on ice when ready - Dilute ascites samples 1:2 in FCS free medium (A2)

- Put diluted ascites in water bath at 37°C, 3-5 minutes before treatment start - Put prepared 6 well plates in water bath that is set to 37°C

- Remove FCS free medium from wells

- Add 0,5 mL (50 %) tempered ascites to each well, start timer

- Put plate on ice immediately after time course (after 20 and 3 minutes)

28

- Remove treatment medium and wash with 1,0 mL/well ice cold PBS (A10) - Remove PBS and add 700 µL/well ice cold Trypsin (A6)

- Trypsinize cells for 20 minutes on ice

- Quench reaction by adding 600 µL /well FCS medium (A3)

- Wash cells with 1 mL pipette and transfer cells to a 1.5 mL Eppendorf tubes - Spin down, 5 min @ 500g @ 4°C (Eppendorf centrifuge - F3)

- Resuspend cells in 1 mL ice cold PBS (A10)

- Spin down, 5 min @ 500g @ 4°C (Eppendorf centrifuge - F3) - Resuspend cells in 100 µL ice cold PBS (A10)

- Add 100 µL FIX buffer I (A14) and incubate in water bath at 37°C for 10 minutes

- Add 800 µL ice cold PBS, spin down for 5 min @ 700g @ 4°C (Eppendorf centrifuge - F3) - Wash cells with 1 mL ice cold PBS (A10)

- Resuspend cells in so that the total volume is 200 µL with ice cold PBS (A10) - Cells are now ready for fluorescent cell barcoding

29

8.2 F

Flow cytometry is a technique that can measure physical cell properties as single cells flow past one or multiple light beams. Scattered and emitted light can be detected and converted to electronic signals. Then computer software can convert these signals into readable data that contains information about physical characteristics of each cell in the sample, including cell size, granularity and relative fluorescent intensity. With flow cytometry a whole spectre of proteins can be investigated simultaneously within single cells, making it possible to draw maps of protein status within individual cells at specific time points. This makes it possible to obtain both qualitative and quantitative information about intracellular molecules and/or cell surface receptors in one single cell.

Other analytical methods like Western blot are usually used for investigating one or a few proteins within a cell population.

In the last decade there has been improvement within flow cytometry and immunostaining methods (staining specific targets with antibodies), which has led to the rise of phosphoflow cytometry. This technique makes it possible to study phosphorylation of intracellular proteins that are essential in signalling transduction within individual cells using phospho-epitope specific antibodies [29].

Furthermore, introduction of fluorescent cell barcoding (FCB) has revolutionized the method by allowing high throughput analysis, making it possible to analyse multiple cell populations in one sample. In the future, phosphoflow cytometry can give us an increased global understanding of the signal transduction dynamics at single cell levels [29].

In our experiments, before analysing samples by phosphoflow cytometry, cell populations were treated (activating signalling transduction). After this, cell populations were barcoded, combined and stained with protein specific antibodies that were conjugated to fluorochromes.

30

Figure 11 Flow cytometry (BDCanto II) Sample flow chamber (A) and side/forward scatter (B), from [30, 31].

In flow cytometry, samples (containing cells) are injected into the middle of the sheath flow (Figure - 11A). If the density and/or velocity of the sample fluid and the sheath fluid differ enough they will not mix (laminar stream), creating a two layer fluid flow. Sheath flow carries cells towards a hydrodynamic focus point where one or multiple lasers are aimed. This focus point is created by increasing or decreasing sample pressure, making it possible to regulate the diameter (Figure 11A).

Increased pressure will provide a higher flow rate and diameter. This means that multiple cells will be exposed to the laser beams at once, leading to a lower resolution [30, 31]. By controlling the flow rate it is possible to acquire a flow diameter where there is only room for one cell, thereby making it possible to measure single cells (Figure 11A).

In BD FACSCanto and BD FACSFortessa there are three lasers; blue laser (488nm), red laser (640nm) and violet laser (405nm). When one or multiple laser beams strike a single cell in the measuring chamber, some light will be scattered and some energy will be absorbed. If fluorescent antibodies are used to stain cells or cellular components, photons are absorbed causing electrons to jump to a higher energy state (electrons are excited). After some time, electrons will return to their ground state, leading to the emission of quantum light, which is known as fluorescence. The emitted wavelength is usually longer than the absorbed wavelength, and this makes it possible to use the same laser on multiple fluorochromes. For example, fluorochromes PerCP and FITC both absorb blue light at 488 nm but PerCP emits red light at 675nm and FITC emits green light at 530nm. Scattered and emitted light passes through a number of lenses and optical filters before reaching the detectors (Figure 12). Their main function is to modify spectral distribution by sorting and guiding specific intervals of wavelength (the emitted light of a specific fluorochrome) to its specific detector.